What technology stack does UBS run, and how does AI sit on it?
UBS's stack is group-wide and built on Microsoft Azure. A 2018 agreement to move a third of applications to public cloud was met early and expanded in October 2022 to more than 50% of applications, including critical workloads, on Azure within five years. Its AI layer sits on that cloud: the Red assistants on Azure OpenAI Service and Azure AI Search over about 60,000 digitized documents, an internal AI Hub and a data mesh (2025); RiskLab, the data-science platform for more than 1,200 data scientists, with an AI Common Ecosystem and a governed PostgreSQL vector store, built on Domino Data Lab; and UBS Claves, the single platform it is consolidating onto, which routes between lightweight and frontier models. Coding is the largest AI workload by headcount: 3,000 developers used AI code tooling in October 2025 and about 18,000 in July 2026, with over a quarter of AI suggestions going straight into production code. Underneath runs the Credit Suisse decommissioning: by April 2026 UBS had retired about 1,700 (60%) of in-scope business applications and switched off 76,000 servers, after ten data-center exits by year-end 2025. UBS Bank USA runs on these group platforms; no separate US stack is disclosed.[1][2][3][4][5][6][7][8][9][10]
Data as of Oct 16, 2026 · UBS profile updated Oct 13, 2026 · 15 sources cited on this page · 140 banks profiled on this site · Markdown
What are the main layers of UBS's technology and AI stack?
As UBS, Microsoft and the trade press describe them. All are group platforms.
| Layer | What it is | Vendor | Date |
|---|---|---|---|
| Public cloud | More than 50% of applications, including critical workloads, to run on Azure within five years; earlier one-third target met ahead of schedule; workloads moved cut energy use by up to 30% | Microsoft Azure | October 2022[1] |
| Knowledge base and assistants (Red) | About 60,000 investment-advice and product documents digitized into a queryable knowledge base; vector search over unstructured data; two domain-specific assistants with multi-language support | Azure AI Search, Azure OpenAI Service | January 2025[3][11] |
| AI Hub and data mesh | An internal AI Hub for exploring new use cases and a data-mesh framework; AI solutions deployed across five divisions | Microsoft Azure | February 2025[2][3] |
| RiskLab and AICE | Data-science platform for 1,200-plus data scientists; the AI Common Ecosystem gives governed production-data access, managed LLM endpoints, tenanted RAG and agent deployment pipelines; model-to-production time cut from six months to one | Azure, MLflow, Azure Kubernetes Service | May 2025[4] |
| VEGA vector store | Governed multi-tenant vector store for RAG use cases, with hybrid, semantic and graph-RAG retrieval | Azure Database for PostgreSQL with pgvector | May 2025[4] |
| Model development and validation | RiskLab built on Domino's platform, to be rolled out across the bank; UBS took an equity stake | Domino Data Lab | August 2025[5] |
| UBS Claves | Single strategic AI platform; lightweight-to-frontier models with automated routing on performance and token cost; evaluation for governance and value tracking | Multiple model providers (not named) | Described 2026[6] |
How far has UBS got with AI coding and retiring Credit Suisse systems?
Group figures from UBS's quarterly results releases.
| Date | AI in engineering | Decommissioning |
|---|---|---|
| July 2025 | Coding tools rolled out with a targeted 15% efficiency gain | About 700 applications decommissioned, 56% of Non-core and Legacy's initial stack[12] |
| October 2025 | 3,000 developers using AI-powered code tooling | 1,365 applications (47%) retired; 66,000 servers switched off; 43PB of data worked through; seven data-center exits[7] |
| February 2026 | Not quantified | 1,598 applications (55%) retired; 71,000 servers off; ten data-center exits[10] |
| April 2026 | Not quantified | About 1,700 business applications (60%) retired; 76,000 servers (71%) switched off[9] |
| July 2026 | About 18,000 developers using AI across code generation, explanation, test automation and requirements; over a quarter of AI suggestions used directly in production code | More than 90% of legacy applications no longer in use, about 70% fully decommissioned[8] |
What does the UBS profile record on this topic?
The rows below are the entries of the UBS AI strategy profile that bear on this page, as of Oct 13, 2026; the profile holds the full record of 28 dated moves, 14 use cases and 35 sources.
| Use case | What the record shows | Status | Source |
|---|---|---|---|
| UBS Claves platform and AI risk committee | One platform with model routing and evaluation; AI Operating & Risk Committee under a group AI policy. | Rolling out | [6] |
| RiskLab model development and validation (Domino) | An auditable AI model-development platform built on Domino Data Lab, being rolled out across the bank. | Rolling out | [5] |
| AI-assisted software development | About 18,000 developers use AI across code generation, explanation, test automation and requirements; over a quarter of AI suggestions go straight into production code. | In production | [8][13] |
| Microsoft Azure OpenAI Service and Azure AI Search | Platform behind the Red assistants and a queryable knowledge base of about 60,000 investment-advice and product documents; Microsoft Azure is named as a strategic AI partner. | In production | [11][6] |
| Date | Move | Detail | Source |
|---|---|---|---|
| Mar 18, 2026 | Credit Suisse client migration completed | UBS says it has moved around 1.2 million former Credit Suisse clients globally and enters the final phase of the integration, on track to substantially complete it by the end of 2026. | [14] |
| Aug 21, 2025 | Equity investment in Domino Data Lab | UBS leads an investment in the provider behind RiskLab, its AI model development and validation platform, which is to be rolled out across the bank. | [5] |
| Dec 31, 2024 | Azure OpenAI Service reaches 30,000 UBS employees | Microsoft says its Azure OpenAI Service and Azure AI Search power UBS's Red assistants and had been rolled out to 30,000 employees as of December 2024, across wealth management, personal and corporate banking and group operations. The page is undated; the figure is as of December 2024. | [11] |
| Jun 1, 2024 | Credit Suisse's US businesses move under one US holding company | UBS AG and Credit Suisse AG merged in May 2024; UBS says the transition to a single US intermediate holding company took place in June 2024, with client and platform migrations still to follow. The day is not published. | [15] |
Which cloud does UBS use?
Microsoft Azure is its strategic public cloud. A 2018 agreement targeted a third of applications in public cloud within four years and was met early; in October 2022 UBS expanded it to more than 50% of applications, including critical workloads, on Azure within five years. UBS has not published a later completion figure.
What is UBS Red built on?
Azure OpenAI Service and Azure AI Search. UBS digitized about 60,000 investment-advice and product documents into a queryable knowledge base, and Red uses vector search over it to answer advisors' questions in several languages. Red was available to more than 85,000 employees by October 2025, according to UBS.
What is UBS RiskLab?
UBS's data-science platform, used by more than 1,200 data scientists across risk, compliance, finance, investment banking and wealth management. Its AI Common Ecosystem provides governed data access, managed LLM endpoints, RAG environments and agent deployment pipelines, with a PostgreSQL vector store called VEGA; it is built with Domino Data Lab.
How many UBS developers use AI coding tools?
About 18,000 as of the second-quarter 2026 results in July 2026, up from 3,000 in October 2025. UBS says they use AI for code generation, explanation, test automation and requirement management, and that over a quarter of AI-generated suggestions are used directly in production code.
Does UBS Bank USA have its own technology stack?
None is disclosed. UBS Bank USA and the US wealth business run on group platforms (Azure, Red, Claves and RiskLab), with US-built analytics such as STAAT on top. The Credit Suisse US businesses moved under a single US holding company in June 2024, and client migration finished globally in March 2026.
- UBS to have 50% of applications running in Microsoft Azure cloud
- UBS and Microsoft partner to advance AI in banking
- UBS underpins AI development with data transformation initiative
- UBS unlocks advanced AI techniques with PostgreSQL on Azure (UBS RiskLab authors)
- UBS Invests in Domino Data Lab, Deepening Strategic Partnership to Drive Next Wave of Enterprise AI Innovation
- Innovation and AI at UBS (undated page, retrieved 2026-10-13)
- UBS reports strong results in 3Q25 with continued progress on integration
- UBS reports 2Q26 net profit of USD 2.8bn and USD 5.8bn for 1H26 (Business Wire)
- UBS reports USD 3.0bn net profit and 16.8% RoCET1 in 1Q26; on track to complete integration by year-end
- UBS reports net profit of USD 1.2bn in 4Q25 and USD 7.8bn in FY25; confirms 2026 targets and sets ambitions for 2028
- UBS and Microsoft unite: co-creating the future of banking with Azure AI (undated; figures as of December 2024)
- UBS reports 2Q25 net profit of USD 2.4bn and USD 4.1bn for 1H25 with integration remaining on track
- UBS Group AG Annual Report 2025 on Form 20-F
- UBS nears full Credit Suisse integration by finishing IT project
- Credit Suisse in the USA (undated page, retrieved 2026-10-13)
When UBS moves on AI, you'll read it here first.
when one of the banks we cover moves on AI, it is in the next morning's brief · six sourced stories · 7 am ET · free
plus every tracker, bank and agent page update, the morning after · leave any morning